Result on logreg l2 benchmark

• platform: Linux4.18.0-305.40.2.el8_4.x86_64-x86_64
• processor: Intel(R) Xeon(R) Gold 6248 CPU @ 2.50GHz
• numpy: 1.21.6 blas=NO_ATLAS_INFO lapack=lapack
• scipy: 1.8.0

Dataset: adult[scaled=True] adult[scaled=False] ijcnn1[scaled=True] ijcnn1[scaled=False] madelon[scaled=True] madelon[scaled=False] news20[scaled=True] news20[scaled=False] rcv1[scaled=True] rcv1[scaled=False]

Objective: L2 Logistic Regression[fit_intercept=True,lmbd=10] L2 Logistic Regression[fit_intercept=False,lmbd=10] L2 Logistic Regression[fit_intercept=True,lmbd=1] L2 Logistic Regression[fit_intercept=False,lmbd=1] L2 Logistic Regression[fit_intercept=True,lmbd=0.1] L2 Logistic Regression[fit_intercept=False,lmbd=0.1] Train loss Train error Test loss Test error value

Chart type: objective curve suboptimality curve relative suboptimality curve bar chart semilog-y semilog-x loglog linear

Quantiles: